Iceland provides a unique opportunity for the study of basalt erosion in a cold oceanic climate ( 5°C to 10°C and 400 to 4000 mm.yr^-1 of precipitation), and for comparing glacier fed-rivers with spring-fed rivers. More than 90% of Iceland consists of basalts, and glaciers take up about 12% of the surface area. The monolithological character allows the signature of lithology to be distinguished from fractionation due to chemical weathering, particularly for the Li isotope fractionation observed in rivers. River water samples show significant variations of (²³⁴U/²³⁸U) (from 1.65 to 1.07) and δ⁶Li for dissolved phases range between -16.8‰ and -23.3‰, suggesting that ⁶Li/⁷Li ratios fractionate during weathering of the source basalts (δ⁶Li-MORB~-4.5‰). High (²³⁴U/²³⁸U) and low δ⁶Li have been found for rivers characterised by low U and Li contents, draining old basalts (>0.7my) in areas with low weathering rates. Low (²³⁴U/²³⁸U) with high δ⁶Li and high U content, have been found in the area highest weathering rates of Icelandic basins.
The correlation of U-series with δ⁶Li strongly supports the hypothesis that both U and Li isotopes reflect the degree of silicate weathering, (Bourdon, et al., 2002).

Temperature fluctuations inferred from δ¹⁸O variations in deep-sea sediment cores and in cores drilled in polar, Greenland and glacier ice, document climate change on sub-Milankovitch time scales of between 10y and 1,000y. The ice cores provide considerable temporal resolution (Grootes, et al., 1993; Mayewski, et al., 1993), but confidence in ages obtained by counting back annual layers may only be 600 y at the 12,500 y level (Alley et al., 1993). The radiocarbon timescale has been recalibrated against U-Th ages on corals back to 30 ky BP (Bard et al., 1993, Burr et al., 1998) but there are still considerable problems obtaining accurate ¹⁴C ages, particularly in calcareous lake sequences because of the 'hard-water' effect (Israelson et al., 1997).

The great advantage of speleothems (carbonate cave deposits) is that they preserve high-resolution records of environmental change, they are protected from erosion, and they yield high precision U-Th dates (e.g. Edwards et al., 1987). Paleoclimate indicators other then the traditional δ¹⁸O-based temperature estimates in stalagmites are being used increasingly. Luminescence variations caused by variations in organic acid concentrations and layer thickness can indicate rainfall, and trapped pollen are indicative of the local flora at the time of calcite deposition.
A major target has been an absolute chronology for the much used biostratigraphic zones in the Late Quaternary. Pollen have been separated from cave deposits that have then been dated by U-Th isotopes as part of a wider project to reconstruct palaeo-environmental changes in the Late Quaternary. New results have established that speleothems were deposited in both interglacial and interstadial periods. For example, a 4 cm flowstone sample from the Mendip Hills is bracketed by ages of 71,000±500 y to 66,300±700 y. Thus, this sample grew rapidly during the early part of the last glaciation, in oxygen isotope stage 4. However, the sample contains pollen of thermophilous species including oak, ash and hazel, suggesting that climatic conditions in the south of England at this time were not as severe as previously supposed for that period. The organic acids in this flowstone also suggest fairly wet soil conditions with little humification of organic matter (McGarry and Baker, in press)

Similarly, material from Stump Cross Caverns in Yorkshire has a basal age of at 118,000 ±1200 y, early in the last interglacial, and pollen that indicate a typical interglacial assemblage of pine, birch, oak, alder, ash, hazel, ivy, grass and herbs. Soil conditions were wet when flowstone growth began and humification levels increased during its growth. When more dates are combined with the pollen and organic acid data, the dated terrestrial record of vegetation change can be extended much further back in time than has previously been possible with radiocarbon dating.

U-series disequilibrium dating of lake sediments is often complicated due to the presence of detrital ²³⁰Th within the sample matrix. The analysis of bulk sample thus provides a ²³⁰Th signature that is comprised of detrital and authigenic ²³⁰Th, and yields an apparent age much

older than reality. Many attempts at leaching such samples have yielded non-reproducible results. However, exciting results from lacustrine sediments from Hawes Water (NW Lancashire, UK) indicate that this particular lake has marls that are less affected by detrital thorium input than many other lakes. Hawes Water, a small shallow lake, is much more responsive to climatic and hydrological change than large deep-water lakes. U-series ages match fairly well with a proxy chronology based on pollen horizons and stable isotopic data, See Figure.
Trace element and stable isotope data for the Holocene sequence at Hawes Water indicate considerable climatic change during the last 10 ky contrary to popular belief that the Holocene in NW Europe since the Younger Dryas as been a climatologically quiet period. This is the first attempt to combine high-resolution stable isotopic and geochemical data with an absolute chronology of a NW European shallow lake.

---